“…Since there was consistent, if not ideal, overall agreement between ELISA, ATLA, and immunoblotting, we felt confident that ELISA was adequate to identify populations with high and low prevalences of serum antibodies to HTLV-I, although not to attempt more sophisticated analyses such as those of household clustering [Kazura et aI., 19871. Gajdusek, 1963;Simmons et al, 1958Gajdusek et al, 1972Tesh et al, 1975Curtain et al, 1965Gajdusek et al, 1972 Adels andGajdusek, 1963;Curtain et al, 1965;Gajdusek et al, 1972 Coker-Vann et al, 198 1 ;Garruto et al, 1974;Simmons et al, 1967 Adels and Gajdusek, 1963;Simmons et al, 1967Curtain et al, 1965 Fig. 2.…”